The problem of the neural encoding of complex auditory patterns in the avian brain will be examined using a rigorous statistical methodology. The research will concentrate on the discrimination task that young male song birds must perform while learning to copy a tutor's song (usually the father's song). The response to biologically relevant auditory stimuli of neurons in three nuclei of the anterior forebrain pathway of the birdsong system will be recorded. A statistical approach will be used to examine which features of the auditory stimuli are being encoded and how these features are encoded within the response patterns of the neurons. A measure of discrimination accuracy, based on principles of information theory, will be calculated for different nuclei in the birdsong system and at different stages of development. This theoretical upper limit of neural discrimination can then be compared to the behavioral performance of the animals. The research will enhance our understanding of the basic principles underlying the development and mode of operation of the neural circuitry involved in complex sensory pattern recognition. This basic knowledge will increase our understanding of the higher functions of the human brain and will help in elucidating the problems underlying certain mental illnesses, in particular those which involve learning deficits.